Avionics system failure prediction based on bacteria evolution and gray neural network

Aiming at the problem of the gray neural network easy to fall into local optimization, a failure prediction algorithm with bacterial evolutionary and gray neural network is proposed. First, a gray neural network fault detection model is established. Then, the bacterial evolutionary algorithm is selected to optimize the initial weights and thresholds of the network to solve the defect of network easy to fall into local optimization. Finally, several kinds of fault prediction effect of the algorithm are compared through the simulation experiment. The simulation results show that the algorithm combining bacterial evolutionary and gray neural network can achieve optimal prediction effect quickly.